Nội dung

Binding of N- and C-terminal anti-prion protein antibodies generates distinct phenotypes of cellular prion proteins (PrPC) obtained from human, sheep, cattle and mouse Thorsten Kuczius1, Jacques Grassi2, Helge Karch1 and Martin H. Groschup3 1 Institute for Hygiene, University Hospital Muenster, Muenster, Germany 2 CEA, Service de Pharmacologie et d’Immunologie, CEA ⁄ Saclay, Gif sur Yvette, France 3 Institute for Novel and Emerging Infectious Diseases, Friedrich Loeffler-Institute, Federal Research Centre for Virus Diseases of Health, Greifswald – Isle of Riems, Germany Keywords antibody; glycotyping; prion protein; PrPC; signal intensity Correspondence T. Kuczius, Institute for Hygiene, University Hospital Münster, Robert Koch Strasse 41, 48149 Münster, Germany Fax: +49 251 9802868 Tel: +49 251 9802897 E-mail: tkuczius@uni-muenster.de Website: http://www.hygiene.uni-muenster.de (Received 14 July 2006, revised 20 December 2006, accepted 12 January 2007) doi:10.1111/j.1742-4658.2007.05691.x Prion diseases are neurodegenerative disorders which cause Creutzfeldt– Jakob disease in humans, scrapie in sheep and bovine spongiform encephalopathy in cattle. The infectious agent is a protease resistant isoform (PrPSc) of a host encoded prion protein (PrPC). PrPSc proteins are characterized according to size and glycoform pattern. We analyzed the glycoform patterns of PrPC obtained from humans, sheep, cattle and mice to find interspecies variability for distinct differentiation among species. To obtain reliable results, the imaging technique was used for measurement of the staining band intensities and reproducible profiles were achieved by many repeated immunoblot analysis. With a set of antibodies, we discovered two distinct patterns which were not species-dependent. One pattern is characterized by high signal intensity for the di-glycosylated isoform using antibodies that bind to the N-terminal region, whereas the other exhibits high intensity for protein bands at the size of the nonglycosylated isoform using antibodies recognizing the C-terminal region. This pattern is the result of an overlap of the nonglycosylated full-length and the glycosylated N-terminal truncated PrPC isoforms. Our data demonstrate the importance of antibody selection in characterization of PrPC. Prion diseases, also known as transmissible spongiform encephalopathies, are a group of neurodegenerative disorders affecting both humans and animals. The human forms encompass sporadic and familiar Creutzfeldt–Jakob disease and the new variant Creutzfeldt–Jakob disease (vCJD), which has been linked to BSE, the bovine spongiform encephalopathy of cattle [1,2]. Scrapie is the prion disease in sheep and goats. The main characteristic of the disease is the accumulation of an abnormal prion protein (PrPSc), thought to be the only infectious agent associated with prion neurodegeneration [3]. The pathogenic mechanism is assumed to involve conversion of physiological cellular prion protein (PrPC) to a pathological isoform (PrPSc) accompanied by a conformational change from a largely a-helical form into a b-sheet structure [4]. In contrast to PrPC, the infectious PrPSc protein is detergent-insoluble. PrPC and PrPSc protein samples can be differentiated by pretreatment with proteinase K (PK), which completely hydrolyses PrPC but only removes 55–70 amino acid residues in the N-terminal region of PrPSc resulting in a molecular reduction of 6–8 kDa. The western blot method is a useful in vitro assay for the characterization of PrPSc and PrPC, in which fully glycosylated mouse PrP migrates at 33–35 kDa Abbreviations BSE, bovine spongiform encephalopathy; PK, proteinase K; PrP, prion protein; SAF, scrapie-associated fibril; vCJD, variant Creutzfeldt–Jakob disease. 1492 FEBS Journal 274 (2007) 1492–1502 ª 2007 The Authors Journal compilation ª 2007 FEBS Glycotyping of PrPC T. Kuczius et al. and the nonglycosylated form at 27 kDa on SDS ⁄ PAGE [5]. PrPSc strains and isolates are distinguished by the size of their PK resistant core protein because differences in the PK cleavage sites in PrPSc have been observed in scrapie, experimental scrapie and ruminant BSE [6,7]. PrPSc exhibit different banding patterns following quantitative immunoblotting by densitometry, which reflects differences in the ratios of the di-, mono- and nonglycosylated PrP. In sporadic cases of human Creutzfeldt–Jakob disease, PrPSc shows a characteristic glycopattern with high signal intensity of the mono-glycosylated isoform which differs from that in ruminant BSE and scrapie PrPSc. In addition to vCJD, the occurrence of other Creutzfeldt– Jakob disease subtypes with differing glycoprofiles and molecular masses has been postulated [8–10]. The ability of prions to cross species barriers is largely dependent on the PrPC sequence homology of the donor and recipient species [11,12]. In addition to species-specific characteristics of PrPSc, there are also notable variations in the glycoform patterns, but the importance of these is not well understood. PrPC is expressed ubiquitously and in a highly conserved form in mammalian species [13,14]. Highest levels were found in neurons and the central nervous system [15,16]. Following expression, PrPC undergoes posttranslational modification involving removal of an N-terminal signal peptide and C-terminal residues in the polypeptide chain and attachment of a glycosylphosphatidylinositol group for cytoplasmic membrane anchorage [17]. The structure is characterized by an N-terminal domain including octapeptide repeats, a central hydrophobic domain and a C-terminal region with two asparagine-linked glycosylation sites and a disulphide bond between cysteine residues [18]. The role of PrPC in cell function is not known, but it has been associated with synaptic, enzymatic and signaling functions, copper binding and transport [19–21]. Copper and heparan sulfate binding have been mediated through its N-terminal domain [22,23]. However, under physiological conditions the N-terminal region can be lost by cleavage [23–28]. From endogenous proteolysis, cleavage sites in human PrPC were mapped at amino acids 110–112 and at residues 80–100 generating N-truncated forms; these are referred to as C1 and C2, respectively. The nonglycosylated forms migrate at 18 and 21–22 kDa, respectively [25]. In the past, little attention was given to the banding patterns and different glycoforms of PrPC. In this study, we have analyzed the glycoform patterns of PrPC of human, sheep, cattle and mice and compared them. Variable immunoreactivity of anti-PrP antibodies determining different PrPC banding patterns is a feature used especially to find heterogeneity based on protein conformation in one species [29]. Independent of individual brain regions, in this study, we focused our analysis on PrPC glycoform patterns derived from different species which arose from binding of various antibodies recognizing sites in the amino, central or C-terminal PrPC sequence. The aim of the study was therefore to find imposing interspecies variations among human PrPC and PrPC derived from different species in order to find a first onset on the basis of PrPC expression why PrPSc of human differed from other species. Using a panel of monoclonal antibodies we systematically analyzed the formed signal intensities of the di-, mono- and nonglycosylated PrPC and the N-truncated isoforms. We found that the mouse PrPC glycoforms differed from human when C-terminal PrP binding antibodies were used. This observation was attributed to the proportion of full-length PrP and the truncated isoform, which was predominant in human, sheep and cattle brains. C-terminal binding antibodies detect full-length nonglycosylated PrPC as well as truncated glycosylated isoforms at the same size. Taken together, first, the banding pattern is largely dependent on the antibody used and, secondly, there are antibodies by which interspecies variations of glycoform ratios are detectable. The findings are important for studies of PrPC function, regulation and expression, as full-length and truncated isoforms of di-, mono- and nonglycosylated proteins are only detectable with antibodies recognizing the C-terminal region and produce altered expression profiles. Results Proteins of brain homogenates derived from different species were separated on SDS ⁄ PAGE and the specific PrPC signals were detected by the western blot technique. The PrPC banding patterns were analyzed using a set of monoclonal antibodies which recognize various epitopes within the prion protein sequence (Fig. 1 and Table 1). The two bands of higher molecular masses are the di- and mono-glycosylated isoforms and the band with the lowest molecular mass is nonglycosylated PrPC. Quantification of the three protein bands was always carried out in the linear range determined using serial dilutions of samples (Fig. 2). Linearity consisting of continuous signal increase and of reproducible glycoprotein patterns was determined in the range between 4 and 10 lL of brain homogenate confirmed by repeated gel runs. Signal intensities therefore were analyzed within continuous, optimal and reproducible glycoform patterns. FEBS Journal 274 (2007) 1492–1502 ª 2007 The Authors Journal compilation ª 2007 FEBS 1493 Glycotyping of PrPC T. Kuczius et al. Fig. 1. Sequence alignment of prion proteins of humans, sheep, cattle and mice. Recognition sites of the antibodies SAF34, P4, 8G8, SAF60 and SAF70 are indicated. Sequences of the species are recognizable by the antibodies are marked in bold letters. Table 1. Monoclonal antibodies for PrP detection. Antibody Isotype Regiona Linear epitope Source Species recognized SAF34 IgG2a 59–89 Hamster scrapie Human, sheep, cattle, mouse P4 IgG1 93–99b Ovine peptide Sheep, cattle 8G8 IgG2a 97–102c Human peptide Human, sheep 6H4 IgG1 144–152 Human peptide Human, sheep, cattle, mouse SAF60 IgG2b 157–161 Hamster scrapie Human, sheep, cattle, mouse SAF70 IgG2b 156–162 Hamster scrapie Human, sheep, cattle, mouse SAF84 IgG2b Octapeptide region (N-terminal region) Intermediary region (N-terminal region) Intermediary region (N-terminal region) Central region (C-terminal region) Central region (C-terminal region) Central region (C-terminal region) Central region (C-terminal region) 126–164d Hamster scrapie Sheep, cattle, mouse a N-terminal region (N terminus; N), C-terminal region (C terminus; C). bLinear epitope of ovine PrP. cLinear epitope of human PrP. dRecognized solid-phase immobilized peptide 126–164, but failed to bind peptide 142–160 [50]. Brain PrPC from humans were detected as two distinct main glycoform patterns, depending on the monoclonal antibody used (Fig. 3). The di-glycosylated isoform was most abundant using antibodies directed against epitopes within the octapeptide or an intermediary region (i.e. amino acids 59–120, designated here as the N-terminal region), but was much less abundant using antibodies binding to the core region (i.e. amino acids 121–166; C-terminal region). The di-glycosylated band of human PrPC showed the heaviest staining with antibodies binding to the N-terminal region (Fig. 3A,B). For example, mAb SAF34, which recognizes the octapeptide sequence, gave a high (50%), an intermediate (29%) and a low (21%) intensity signal with di-, mono- and nonglycosylated PrPC, respectively. Similar ratios were obtained with human PrPC and the antibody mAb 8G8 which binds to the intermediary region at amino acids 97–102 of human PrPC (Fig. 3A,B). In contrast, deviant profiles were found with antibodies binding to the central region of PrPC, 1494 as the signal intensities at the size of the nonglycosylated full-length PrPC (at 27 kDa) were high with monoclonal antibodies 6H4, SAF60 and SAF70 while signals for the di-glycosylated PrPC were low. In these experiments the mono-glycosylated forms of human PrPC were almost invisible and not detectable. Heterogeneity of PrPC proteins is enhanced by endogenous proteolytic modifications, which occurs in vivo [25–27]. PrPC from non-infected brains consists in addition to full-length PrP to a significant amount of an N-terminal truncated PrPC fragment termed C1. Glycosylated C1 protein fragments migrate to a position around the nonglycosylated full-length PrPC. The degree of truncated PrPC to full-length PrP was analyzed after deglycosylation. While N-terminal binding antibodies as SAF34 detected only full-length PrPC, C-terminal binding antibodies recognized two bands comprising full-length PrPC and an 18–19 kDa protein band corresponding to the N-terminally truncated form. The distribution of the signal intensities of the FEBS Journal 274 (2007) 1492–1502 ª 2007 The Authors Journal compilation ª 2007 FEBS Glycotyping of PrPC T. Kuczius et al. A µl kDa 36 0.5 1 2 4 6 8 10 12 27 B 10000000 1000000 100000 units 10000 1000 100 10 1 0 C glycosylation (%) Fig. 2. Western blot analysis and determination of the linear range for signal increase and consistently reproducible glycoprotein banding patterns. (A) Immunodetection of PrPC derived from pooled cattle brain homogenates (10%; 0.5, 1.0, 2.0, 4.0, 6.0, 8.0, 10.0 and 12.0 lL). Antibody p4 was used for detection. (B) PrP proteins were measured by densitometry and quantified using QUANTITY ONE software. The combined PrP signals are given as computer internal units to determine the linear range of reaction. (C) For glycotyping, the combined PrP signals for the di- (d), mono- (j) and nonglycosylated (m) isoform were defined as 100% and the contribution of each band was calculated as percentage. Linearity in the range of 4–10 lL of brain homogenates was confirmed by repeated separate gel runs. 2 4 6 8 homogenate suspension (µl) 10 12 100 80 60 40 20 0 0 two bands demonstrated a higher intensity of the C1 fragment than intensity of full-length PrPC. Human C1 fragments revealed high signal intensities with antibodies SAF60, SAF70 and 6H4 compared with full-length PrPC (Fig. 3C). Observations similar to human PrPC were also observed with PrPC from cattle, sheep and mice. High signal ratios were determined for di-glycosylated ovine PrPC with antibodies SAF34, 8G8 and P4 and lower intensities for mono- and nonglycosylated ovine PrPC (Fig. 4A,B). However, antibodies 6H4, SAF60, SAF70 and SAF84 gave rather low signal intensities for the di-glycosylated isoforms and highest intensities for proteins at 27 kDa, which comprise full-length nonglycosylated PrPC and glycosylated N-terminal truncated isoforms. However, the intensity of the monoglycosylated band was not dependent on the choice of antibody. After deglycosylation, high signal intensity was determined for the truncated isoform and low intensity of deglycosylated full-length proteins (Fig. 4C). 2 4 6 8 homogenate suspension (µl) 10 12 Results similar to these were obtained with PrPC from cattle where the antibodies SAF34 and P4 strongly stained the di-glycosylated band, and mAbs 6H4, SAF60, SAF70 and SAF84 showed the highest staining with the overlapping bands of nonglycosylated full-length PrPC and glycosylated truncated isoforms (Fig. 5A–C). In the case of murine PrPC, N-terminal antibodies showed less pronounced staining with the di-glycosylated PrPC than those recognizing the central region. Antibodies 6H4, SAF60, SAF70 and SAF84 gave strong signals for full-length PrPC and less intense signals for the truncated fragments (Fig. 6A–C). Taken together, these findings indicate that the signal intensities of PrPC glycoform patterns strongly depend on the choice of the antibody which was used and to a lesser extent on the species from which the PrPC was obtained (Fig. 7). The di-glycosylated PrP protein bands of humans, sheep, cattle and mice were always predominant, with antibodies binding to the N-terminal region. These patterns changed when PrPC FEBS Journal 274 (2007) 1492–1502 ª 2007 The Authors Journal compilation ª 2007 FEBS 1495 Glycotyping of PrPC A T. Kuczius et al. A human PrP binding antibodies N-terminal C-terminal sheep PrP binding antibodies N-terminal C-terminal kDa kDa 36 36 27 27 SAF34 8G8 6H4 SAF60 SAF34 8G8 SAF70 P4 6H4 SAF60 SAF70 SAF84 B B 80 glycosylation (%) glycosylation (%) 100 80 60 40 20 40 20 0 0 SAF34 8G8 6H4 SAF60 SAF34 8G8 SAF70 P4 6H4 SAF60 SAF70 SAF84 C C kDa kDa 27 27 20 20 SAF34 6H4 SAF60 SAF70 Fig. 3. (A) Western blot analysis of human PrPC. Proteins of brain homogenates were separated by SDS ⁄ PAGE followed by immunoblotting. PrPC signals were detected using the antibodies indicated. (B) The glycoforms of the protein bands were analyzed by calculation of the percentages of the di- (d), mono- (j) and nonglycosylated (m) isoform as arithmetic means of separate gel runs. The number of gel runs for the analyses are given for each antibody. Accounting for differences among gel runs, SE values were calculated according to antibody used for PrP detection. Calculation of the banding patterns of 10 gels using antibody SAF34 gave an SE value of 2.1 for the di-glycosylated isoform, 1.6 for the mono-glycosylated band and 3.2 for the nonglycosylated protein; six gels using antibody 8G8 (SE 1.1; 0.6; 1.1); seven gels with antibody 6H4 (SE 0.4; 0; 0,4); seven gels with antibody SAF60 (SE 1.2; 0; 1.2); and 13 gels with antibody SAF70 (SE 0.8; 0.4; 1.0). (C) Electrophoretic pattern of deglycosylated PrPC. Brain homogenates were treated with PNGase F and proteins were separated by SDS ⁄ PAGE. Deglycosylated full-length PrPC and the N-terminal truncated forms (C1) were detected using the antibodies indicated. Results were confirmed by repeated separate gel runs per antibody. was detected by C-terminal binding antibodies. A protein band with highest signal intensity at the size of the nonglycosylated PrPC was determined for humans, sheep and cattle. This high signal intensity resulted 1496 60 SAF34 6H4 SAF60 SAF70 SAF84 Fig. 4. (A) Immunoblotting of proteins derived from sheep brain homogenates. PrPC signals were specifically detected using the antibodies indicated. (B) Signal intensities of the di- (d), mono- (j) and nonglycosylated isoform (m) of PrPC were quantified and calculated as percentages of the total signal. The glycoforms of the protein bands were analyzed as arithmetic means of separate gel runs. The number of gel runs are given for each antibody, and, accounting for differences among gel runs, SE values were calculated according to antibody used for PrP detection. Calculation of the banding patterns of 17 gels using antibody SAF34 gave an SE of 2.1 for the di-glycosylated isoform, 1.0 for the mono-glycosylated band and 1.6 for the nonglycosylated protein; five gels using antibody 8G8 (SE 1.5; 1.1; 0.5); 30 gels with antibody P4 (SE 0.9; 0.7; 1.2); five gels with antibody 6H4 (SE 4.6; 3.9; 4.7); seven gels with antibody SAF60 (SE 1.1; 1.0; 1.4); 30 gels with antibody SAF70 (SE 1.9; 1.9; 2.9) and nine gels with antibody SAF84 (SE 1.8; 2.3; 3.4). (C) Brain homogenates were treated with PNGase F for deglycosylation of the proteins and subjected to immunoblotting. Full length PrPC and the N-terminal truncated forms were detected using antibodies indicated. The proportion of full-length PrPC and truncated isoforms was confirmed by repeated separate gel runs. from an overlay of nonglycosylated full-length and glycosylated truncated PrPC. However, mouse PrPC, in most cases, showed highest intensities for the di-glycos- FEBS Journal 274 (2007) 1492–1502 ª 2007 The Authors Journal compilation ª 2007 FEBS Glycotyping of PrPC T. Kuczius et al. A cattle PrP binding antibodies N-terminal C-terminal A mouse PrP binding antibodies N-terminal C-terminal kDa kDa 36 36 27 27 SAF34 P4 6H4 SAF60 SAF70 SAF84 B B 60 40 20 0 SAF34 P4 6H4 SAF34 6H4 SAF60 SAF70 SAF84 60 40 20 0 SAF60 SAF70 SAF84 C kDa 6H4 80 glycosylation (%) glycosylation (%) 80 SAF34 SAF60 SAF70 SAF84 SAF70 SAF84 C kDa 27 27 20 SAF34 6H4 20 SAF60 SAF70 SAF84 SAF34 Fig. 5. (A) Western blot analysis of brain tissues obtained from cattle. After immunoblotting, PrPC signals were detected using the antibodies indicated. (B) The protein banding pattern of the three PrPC protein bands, the di- (d), mono- (j) and nonglycosylated isoform (m), was analyzed using densitometry. The percentages of each band regarding to the total signal of PrPC were calculated as arithmetic means of separate gel runs. The number of gel runs for the analyses are given for each antibody. Considering differences among gel runs, SE values were calculated according to antibody used for PrP detection. Calculation of the banding patterns of six gels using antibody SAF34 gave an SE of 1.7 for the di-glycosylated isoform, 0.3 for the mono-glycosylated band and 0.9 for the nonglycosylated protein; 17 gels using antibody P4 (SE 1.3; 1.2; 0.5); six gels with antibody 6H4 (SE 1.6; 1.0; 0.9); 13 gels with antibody SAF60 (SE 3.1; 1.5; 3.1); 21 gels with antibody SAF70 (SE 1.1; 1.1; 1.3); and eight gels with antibody SAF84 (SE 1.1; 1.7; 1.1). (C) Proteins of cattle brain homogenates were deglycosylated using PNGase F followed by immunoblotting. Signals of full-length PrPC and truncated PrPC were detected using the antibodies indicated and the patterns were confirmed by repeated gel runs. ylated band. A differentiation of mouse PrPC to other species is feasible by antibodies recognizing the C-terminal region. The comparison of PrPC patterns from brains of humans, sheep, cattle and mouse demonstrated consistent differences in the proportion of the C1 fragment. 6H4 SAF60 Fig. 6. (A) Detection of mouse PrPC by western blotting. Proteins of brain homogenates were immunoblotted and PrPC signals were detected using the antibodies indicated. (B) Signals of each of the three PrPC protein bands, the di- (d), mono- (j) and nonglycosylated isoform (m), were quantified. The number of gel runs for the analyses are given for each antibody. Following differences among gel runs, many gel runs were analyzed. The percentages of the PrPC bands were calculated as arithmetic means and SE according to the antibody used for PrP detection. Calculation of the banding patterns of 16 gels using antibody SAF34 gave an SE of 1.2 for the di-glycosylated isoform, 0.9 for the mono-glycosylated band and 0.5 for the nonglycosylated protein; six gels with antibody 6H4 (SE 1.9; 1.1; 1.1); four gels with antibody SAF60 (SE 0.7; 0.6; 0.7); 17 gels with antibody SAF70 (SE 1.4; 0.6; 1.6); and nine gels with antibody SAF84 (SE 2.5; 1.0; 2.1). (C) Brain homogenates were treated with PNGase F. After immunoblotting, membranes were probed with the antibodies indicated. Repeated gel runs confirmed the proportion of full-length and truncated PrPC. According to these results, PrPC banding patterns seem to depend strongly on the choice of the antibody used for detection and also, albeit to a lesser extent, on the species of origin from which PrPC derived. As PrPSc and PrPC glycoform patterns in humans have previously been reported to vary considerably in the FEBS Journal 274 (2007) 1492–1502 ª 2007 The Authors Journal compilation ª 2007 FEBS 1497 Glycotyping of PrPC T. Kuczius et al. human sheep mouse cattle 100 80 60 % 40 20 0 N- C- N- N- C- C- N- C- -terminal binding antibodies Fig. 7. Comparison of the PrPC banding patterns of various species detected by amino- and carboxyl-binding antibodies. After immunoblotting, PrPC proteins were detected using N- or C-terminal binding antibodies. The signal intensity of each of the three protein bands was quantified by densitometry. The mean values of the calculated signal intensities were analyzed for each of the N- or C-terminal binding antibodies. The banding pattern of the di- (d), mono- (j) and nonglycosylated isoform (m) is shown for human, sheep, cattle and mouse. The calculation is composed of signals from the N-terminal binding antibodies SAF34, P4 and 8G8 or the C-terminal binding antibodies 6H4, SAF60, SAF70 and SAF84 in consideration of species recognition. Values are calculated for the N-terminal antibodies SAF34 and 8G8 for humans, SAF34, 8G8 and P4 for sheep, SAF34 and P4 for cattle, and SAF34 for mice; and for the C-terminal antibodies 6H4, SAF60 and SAF70 for humans, and mAbs 6H4, SAF60, SAF70 and SAF84 for sheep, cattle and mouse. A kD a 36 27 c cb bs B 100 75 % 50 25 0 c same individual depending on the kind of tissue samples that were analyzed and even between different brain regions [29], we have examined whether this is also reflected in the PrPC glycoform patterns of ovine PrPC which originated from different brain regions such as cortex, cerebellum and brain stem. To give evidence that the banding profile is mostly the result of the antibody recognizing the N- or C-terminal PrP sequence, we analyzed three different brain regions pooled from three individual sheep. Interestingly, we found only small regional independent differences on the antibody used (Fig. 8A.B). Only brain stem seems to contain a slightly smaller di-glycosylated PrPC fraction as compared with that found in the two other regions. However, a major antibody-associated effect was once again observed for PrPC glycoprotein patterns for all three regions: di-glycosylated PrPC bands were heavily stained by N-terminally binding antibody SAF34. Lower intensities were recorded for monothan for nonglycosylated PrPC. However, the glycoform pattern was remarkably different again, when PrPC was detected by SAF70: there was a high signal intensity of proteins at the size of full-length nonglycosylated PrPC, a low intensity for the di-glycosylated isoform and the mono-glycosylated isoform was only just undetectable. A protein band at 27 kDa was most abundant, resulting in an overlay of the full-length cb bs C P NG as e F k Da 27 c + cb + bs + + + + 20 SAF34 (N-terminal binding antibody) k Da 36 27 100 75 % 50 25 0 PNG a se F kD a 27 20 SAF70 (C-terminal binding antibody) Fig. 8. Immunoblot analysis and diagrammatic presentation of PrPC bands obtained from three different regions of sheep brains. (A) Immunodetection of PrPC derived from cortex (c), cerebellum (cb) and brain stem (bs) of sheep detected by antibodies SAF34 and SAF70, respectively. (B) PrPC signals of cortex (c), cerebellum (cb) and brain stem (bs) were quantified and the percentages of the di- (d), mono- (j) and nonglycosylated isoform (m) were calculated as arithmetic means of separate gel runs. The calculation represents seven, nine and nine gels for cortex, cerebellum and brain stem samples, respectively, detected by SAF34, and nine gels each for the different regions detected by SAF70. To account for differences among gel runs, SE values were calculated. SE values of PrPC of cerebrum, cerebellum and brain stem detected by SAF34 were determined for the di- (3.5; 2.1; 1.2), mono- (1.4; 0.9; 0.5) and nonglycosylated isoform (2.2; 1.5; 1.0); and detected by SAF70 were determined for the di- (1.8; 3.1; 1.2), the mono- (0.3; 3.3; 3.6) and the nonglycosylated isoforms (1.8; 5.8; 2.8), respectively. (C) Deglycosylation of PrPC from cortex (c), cerebellum (cb) and brain stem (bs). Proteins were incubated with PNGase F before electrophoresis and transfer to membranes. PrP proteins were detected using antibodies SAF34 or SAF70 as indicated. 1498 FEBS Journal 274 (2007) 1492–1502 ª 2007 The Authors Journal compilation ª 2007 FEBS Glycotyping of PrPC T. Kuczius et al. nonglycosylated PrPC and the N-terminal truncated isoform shown after deglycosylation (Fig. 8C). The truncated C1 fragment exhibited higher signal intensity than the full-length PrPC, indicating a predominance of the truncated isoforms in cortex, cerebellum and brain stem. Discussion The western blotting technique is frequently used for the diagnostic confirmation of prion diseases and to distinguish between the various prion strains. However, the sensitivity of PrPSc to treatment with PK and the glycotyping pattern obtained depend on the prion strain [1,6–9,30–35]. PK treatment reflects in the molecular mass of the initial PK-resistant cleavage product and the reaction kinetics under high proteolytic conditions. The PK cleavage sites have been shown to differ between species, e.g. residue N96 (and Q97 as minor site) in PrPSc from BSE while in scrapie, cleavage is at G81, G85 and G89 (or mainly G89 under different PK concentrations) [36]. In different cases of Creutzfeldt–Jakob disease, two primary cleavage sites at residues 82 and 97 for types 1 and 2, respectively, have been identified; minor cleavage points are present at residues 74–102 [37]. Differences in the glycoprotein pattern are due to differences in the relative staining intensities of the di-, mono- and nonglycosylated isoforms of PrPSc. BSE and human vCJD, the latter presumably being linked to the consumption of BSE-contaminated meat, have a similar glycoprotein profile [1] that can be distinguished from that found in sporadic Creutzfeldt–Jakob disease and sheep scrapie. PrPC serves as the substrate for the PrPSc conversion reaction. However, little is known about the glycoprotein patterns found in PrPC of animal and human origin, and about the effect which the detection antibody might have on these. Brain regional variability of PrPC has been described [29,38]. We systematically analyzed the PrPC glycoform patterns in human, sheep, cattle and mouse brains using a set of antibodies recognizing several epitopes within various regions of the PrP sequence in order to find imposed interspecies variations. Irrespective of the species and of pooled sheep brain regions analyzed, two representative PrPC glycoform patterns were observed depending on the antibody used. Antibodies to the nonstructured N-terminus gave significantly stronger signals with the di-glycosylated isoform of PrPC than did antibodies to the structured core region. However, the glycoform patterns of mouse PrPC always showed the highest signal intensity of the di-glycosylated isoform, independ- ently if an N- or C-terminal binding antibody was used. In contrast, a protein band at the size of the nonglycosylated full-length PrPC of humans, sheep and cattle was highly abundant when using C-terminal binding antibodies. Our data show that the high signal intensity corresponding to the size of the nonglycosylated full-length protein indicated antibody binding at the structured core region of PrPC as the result of an overlap of two proteins, the nonglycosylated full-length form and the glycosylated N-truncated fragments. From endogenous proteolysis, two amino truncated isoforms termed C1 and C2 are described migrating at 18 and 21–22 kDa with human PrPC, respectively [23–28]. A separation of both protein isoforms, full-length and N-truncated, could clearly be demonstrated after enzymatic deglycosylation. Interestingly, truncated C1 fragments of human, sheep and cattle PrPC resulted in higher signal intensities than their full-length proteins. However, this observation is different to the mouse PrPC banding pattern. On the basis of differences in the proportions of the signal intensities of full-length and truncated isoforms, we suggest that PrPC metabolism and regulation varies among the different species. The N-terminal cleavage of PrPC in vivo may be the result of a downregulation of functions arranged by the N-terminal region [26]. The occurrence of two distinct glycoform patterns demonstrated by antibodies binding to the N- or C-terminal region is most likely to be due to differences in epitope and protein fragment accessibility rather than to differences in the glycosylation of PrPC. As shown by NMR (13C, 15N, 1H) and ⁄ or X-ray studies, PrPC in all species contains a flexible N-terminus (amino acids 23–120) [39–41] and a structured core and C-terminal region (amino acids 121– 231). This folded domain contains three helices and two short antiparallel b-sheets [41]. PrPC has two linked glycosylation sites at asparagines 180 and 196 (calculated here for murine PrP) [18]. Taken together, the results of various signal intensities of the three PrPC bands are accredited to the development of the truncated isoforms, to the epitope recognition of the antibodies and in part to the protein structure. These data illustrate that emergent truncated fragments must be taken into account when studying the expression and regulation of PrPC in consideration of the di-, mono- and nonglycosylated protein bands. For distinct discrimination among various species, such as mouse, sheep, cattle and humans, C-terminal binding antibodies will provide more detailed variations in PrPC glycoprotein patterns than antibodies recognizing the N-terminal PrP region. FEBS Journal 274 (2007) 1492–1502 ª 2007 The Authors Journal compilation ª 2007 FEBS 1499 Glycotyping of PrPC T. Kuczius et al. Experimental procedures Antibodies The monoclonal Ig61, Ig62a and Ig62b antibodies (mAbs) used in this study, SAF34, SAF60, SAF70 and SAF84, have been raised in PrP ⁄  mice by immunizing with formic acid-denatured, SAF obtained from an infected hamster brain (263K) [42]. The linear epitopes recognized by these antibodies were identified by pepscan analysis as described [43]. All antibodies were applied as ascetic fluids obtained in mice and used in this study from one charge in each case. mAbs 8G8 and 6H4 (Prionics, Schlieren, Switzerland) were raised against recombinant human PrP [44–46]. A synthetic peptide based on the amino acid sequence of ovine PrP (amino acids 89–104) was used as antigen for producing the monoclonal antibody P4 (r-biopharm, Darmstadt, Germany) [47]. Pepscan analysis revealed P4 peptides at the sequence 93–99 of ovine PrP [48]. The epitopes recognized by the various antibodies and the detection of PrPC derived from various species are listed in Table 1. Preparation of brain tissue Brain tissue was obtained from noninfected sheep, cattle, mice and humans. Homogenates of mice were prepared using pooled whole brains from four individuals. Human homogenates derived from pooled tissues obtained from several different brain regions of six subjects. The regions were not specified, but were comprised mostly of cortex and cerebellum. Brain homogenates of cattle were obtained from the brain stems of six animals. Pooled homogenates of sheep brains were prepared from tissues taken from various regions of five animals. Furthermore, based on three individual sheep, brain tissues of cortex, cerebellum and brain stem were each pooled. The homogenates were prepared by homogenization in nine volumes of lysis buffer [0.32 m sucrose, 0.5% (w ⁄ v) igepal and 0.5% (w ⁄ v) SDS in Tris-buffered saline (20 mm Tris and 150 mm NaCl, pH 7.4; Sigma, Taufkirchen, Germany)] in glass homogenizers followed by intensive ultrasonification as described [49]. After centrifugation at 900 g for 5 min (5415 R centrifuge, FA-45-24-11 rotor, Eppendorf, Hamburg, Germany), the supernatants were stored in aliquots at )70 C. Aliquots mixed with SDS loading buffer were stored at )20 C and were used within a few days in order to avoid effects of prolonged storage on the stability of PrPC. Deglycosylation For enzymatic deglycosylation, SDS was added to the homogenates to a final concentration of 1.5% (w ⁄ v). The protein samples were diluted 2.5-fold in incubation buffer consisting of Tris-buffered saline (20 mm Tris and 150 mm 1500 NaCl; pH 7.4) with 10 mm EDTA, 1% (w ⁄ v) igepal and 1.5% (v ⁄ v) 2-mercaptoethanol. Protein samples were denatured at 99 C for 10 min followed by incubation with one unit of N-glycosidase F (PNGase F; Roche, Mannheim, Germany) for 16 h at 37 C. Non-deglycosylated samples were treated in the same way, but were incubated without the addition of PNGase F. Finally, SDS-loading buffer was applied to the samples processing for SDS ⁄ PAGE. Immunoblot analysis Proteins were separated using SDS ⁄ PAGE. Samples were resuspended in SDS-loading buffer, heated to 99 C for 5 min and the proteins separated in a mini slab gel apparatus (BioRad, Munich, Germany) using 13% polyacrylamide gels. After electroblotting onto Immobilon-P membranes (Roth, Karlsruhe, Germany) using a semi-dry blotting system (Roth), membranes were blocked in Tris-buffered saline containing 0.1% (w ⁄ w) Tween 20 (TBST) and 1% (w ⁄ v) nonfat dry milk powder for 60 min. Specific binding of antibodies to PrP proteins was determined by incubating membranes for at least 2 h with the antibodies indicated. Horseradish peroxidase-conjugated affinity purified goat (anti-mouse IgG) (Dianova, Hamburg, Germany) served as secondary antibody. Protein signals were visualized using a chemiluminescence enhancement kit (Pierce, Bonn, Germany). Glycotyping of prion proteins In order to analyze the PrP glycoform patterns, proteins were scanned on a chemiluminescence photo-imager (BioRad, Munich, Germany). Densitometry was carried out using quantity one software (Bio-Rad, Munich, Germany), determining the signal intensities of the di-, monoand nonglycosylated PrP isoforms. The combined signals with one sample were defined as 100% and each band was calculated as a percentage of the total signal. Protein profiles were analyzed by calculation of the arithmetic means of the tissue samples after separation on SDS ⁄ PAGE. Variations in separation in repeat SDS ⁄ PAGE runs were expressed as standard errors of the mean (se). Acknowledgements The authors thank O. Mantel and O. Böhler for their excellent technical assistance. We are indebted to K. Keyvani, Institute for Neuropathology, Münster, for providing human brain samples, the Chemisches Landes- und Staatliches Veterinäruntersuchungsamt (CVUA) Münster for providing sheep and cattle samples and the Max Planck Institute, Department Vascular Cell Biology, Münster, for providing mouse samples. This work was supported in part by grants from the EU Network Neuroprion (FOOD-CT-2004– FEBS Journal 274 (2007) 1492–1502 ª 2007 The Authors Journal compilation ª 2007 FEBS Glycotyping of PrPC T. Kuczius et al. 506579) and the Bundesministerium für Bildung und Forschung (BMBF; project 0312733). References 1 Collinge J, Sidle KC, Meads J, Ironside J & Hill AF (1996) Molecular analysis of prion strain variation and the aetiology of ‘new variant’ CJD. Nature 383, 685–690. 2 Bruce ME, Will RG, Ironside JW, McConnell I, Drummond D, Suttie A, McCardle L, Chree A, Hope J et al. (1997) Transmissions to mice indicate that ‘new variant’ CJD is caused by the BSE agent. Nature 389, 498–501. 3 Prusiner SB (1991) Molecular biology of prion diseases. Science 252, 1515–1522. 4 Pan KM, Baldwin M, Nguyen J, Gasset M, Serban A, Groth D, Mehlhorn I, Huang Z, Fletterick RJ, Cohen FE et al. (1993) Conversion of alpha-helices into betasheets features in the formation of the scrapie prion proteins. Proc Natl Acad Sci USA 90, 10962–10966. 5 Haraguchi T, Fisher S, Olofsson S, Endo T, Groth D, Tarentino A, Borchelt DR, Teplow D, Hood L, Burlingame A et al. (1989) Asparagine-linked glycosylation of the scrapie and cellular prion proteins. Arch Biochem Biophys 274, 1–13. 6 Kuczius T & Groschup MH (1999) Differences in proteinase K resistance and neuronal deposition of abnormal prion proteins characterize bovine spongiform encephalopathy (BSE) and scrapie strains. Mol Med 5, 406–418. 7 Groschup MH, Kuczius T, Junghans F, Sweeney T, Bodemer W & Buschmann A (2000) Characterization of BSE and scrapie strains ⁄ isolates. Arch Virol Suppl 16, 217–226. 8 Gambetti P, Kong Q, Zou W, Parchi P & Chen SG (2003) Sporadic and familial CJD: classification and characterisation. Br Med Bull 66, 213–239. 9 Hill AF, Joiner S, Wadsworth JD, Sidle KC, Bell JE, Budka H, Ironside JW & Collinge J (2003) Molecular classification of sporadic Creutzfeldt-Jakob disease. Brain 126, 1333–1346. 10 Notari S, Capellari S, Giese A, Westner I, Baruzzi A, Ghetti B, Gambetti P, Kretzschmar HA & Parchi P (2004) Effects of different experimental conditions on the PrPSc core generated by protease digestion: implications for strain typing and molecular classification of CJD. J Biol Chem 279, 16797–16804. 11 Chen SG & Gambetti P (2002) A journey through the species barrier. Neuron 34, 854–856. 12 Kocisko DA, Priola SA, Raymond GJ, Chesebro B, Lansbury PT Jr & Caughey B (1995) Species specificity in the cell-free conversion of prion protein to proteaseresistant forms: a model for the scrapie species barrier. Proc Natl Acad Sci USA 92, 3923–3927. 13 Oesch B, Westaway D & Prusiner SB (1991) Prion protein genes: evolutionary and functional aspects. Curr Top Microbiol Immunol 172, 109–124. 14 Wopfner F, Weidenhofer G, Schneider R, von Brunn A, Gilch S, Schwarz TF, Werner T & Schatzl HM (1999) Analysis of 27 mammalian and 9 avian PrPs reveals high conservation of flexible regions of the prion protein. J Mol Biol 289, 1163–1178. 15 Bendheim PE, Brown HR, Rudelli RD, Scala LJ, Goller NL, Wen GY, Kascsak RJ, Cashman NR & Bolton DC (1992) Nearly ubiquitous tissue distribution of the scrapie agent precursor protein. Neurology 42, 149–156. 16 Kretzschmar HA, Prusiner SB, Stowring LE & DeArmond SJ (1986) Scrapie prion proteins are synthesized in neurons. Am J Pathol 122, 1–5. 17 Stahl N, Baldwin MA, Burlingame AL & Prusiner SB (1990) Identification of glycoinositol phospholipid linked and truncated forms of the scrapie prion protein. Biochem 29, 8879–8884. 18 Caughey B (1993) Scrapie associated PrP accumulation and its prevention: insights from cell culture. Br Med Bull 49, 860–872. 19 Lasmezas CI (2003) Putative functions of PrPc. Br Med Bull 66, 61–70. 20 Maglio LE, Perez MF, Martins VR, Brentani RR & Ramirez OA (2004) Hippocampal synaptic plasticity in mice devoid of cellular prion protein. Brain Res Mol Brain Res 131, 58–64. 21 Mouillet-Richard S, Ermonval M, Chebassier C, Laplanche JL, Lehmann S, Launay JM & Kellermann O (2000) Signal transduction through prion protein. Science 289, 1925–1928. 22 Brown DR, Qin K, Herms JW, Madlung A, Manson J, Strome R, Fraser PE, Kruck T, von Bohlen A, SchulzSchaeffer W et al. (1997) The cellular prion protein binds copper in vivo. Nature 390, 684–687. 23 Chen SG, Teplow DB, Parchi P, Teller JK, Gambetti P & Autilio-Gambetti L (1995) Truncated forms of the human prion protein in normal brain and in prion diseases. J Biol Chem 270, 19173–19780. 24 Shyng SL, Huber MT & Harris DA (1993) A prion protein cycles between the cell surface and an endocytic compartment in cultured neuroblastoma cells. J Biol Chem 268, 15922–15928. 25 Jimenez-Huete A, Lievens PM, Vidal R, Piccardo P, Ghetti B, Tagliavini F, Frangione B & Prelli F (1998) Endogenous proteolytic cleavage of normal and diseaseassociated isoforms of the human prion protein in neural and non-neural tissues. Am J Pathol 153, 1561– 1572. 26 Laffont-Proust I, Hassig R, Haik S, Simon S, Grassi J, Fonta C, Faucheux BA & Moya KL (2006) Truncated PrPc in mammalian brain: interspecies variation and location in membrane rafts. Biol Chem 387, 297–300. 27 Nieznanski K, Rutkowski M, Dominik M & Stepkowski D (2005) Proteolytic processing and glycosylation influence formation of porcine prion protein complexes. Biochem J 387, 93–100. FEBS Journal 274 (2007) 1492–1502 ª 2007 The Authors Journal compilation ª 2007 FEBS 1501